Don't Ignore the Qualitative, Stupid! Why Qualitative Data Matters in Simulations Event

Programme:

13:00 – 13:30
Welcome coffee/tea + registration
13:30 – 14:15
The importance of taking qualitative data seriously in simulation (Professor Bruce Edmonds)
14:15 – 14:45

• A survey on the use of qualitative data to inform Agent-Based Models (Christopher Frantz et al.)
• A review on the use of games to inform Agent-Based Models (Timo Szczepanska).

14:45 – 15:00
Coffee/Tea break
15:00 – 15:30

• How qualitative data is captured by the RAT protocol on rigorous and transparent documentation of data use (Sebastian Archer et al.)
• Awaiting title (Patrycja Antosz)

15:30 – 16:15
“Panel: Don't Ignore the Qualitative, Stupid! Why Qualitative Data Matters in Simulations” (Bruce Edmonds, Juliet Rouchier, Martin Kunc).

Abstracts

The importance of taking qualitative data seriously in simulation (Professor Bruce Edmonds)

Qualitative data is a valuable source of evidence and should not be ignored. As with all kinds of data, textual data poses particular problems for the simulator - in particular, (a) the difficulties of relating it to formal models and (b) the difficulties of interpretation. The former problem has been made much easier with the advent of agent-based modelling, which are open to the implementation of heterogeneous, micro-level narratives of what people do, but which can later be checked against other data (e.g. macro-level time-series). The difficulties of interpretation are being tackled by a range of techniques that are being developed to make the process of interpretation more transparent so they can be critiqued and checked. Two classic examples of simulation work that utilises qualitative data are discussed Bharwani's 2004 work on decision making by farmers, and Taylor's 2003 simulation work on the introduction of e-commerce in supply chains. Qualitative data is not easy - it confronts us with some of the complication and sheer "mess" involved in social systems. We have the choice between deliberately ignoring this and finding rigorous ways of dealing with it. The former is more comfortable, more constrained, and familiar but ultimately less scientific.

Speaker Information

Bruce Edmonds is Professor of social simulation and Director of the Centre for Policy Modelling (CfPM) at the Manchester Metropolitan University. He has been doing agent-based social simulation since 1994 and, with its founder Scott Moss, helped develop the CfPM into one of the world's leading research centres in this field. Together with his colleague, Ruth Meyer, he edits the authoritative handbook in the field, "Simulating Social Complexity", which is in its second edition (2017). He is co-organiser of the European Social Simulation Association SIG on "Using Qualitative Evidence to Specify Social Simulation" with Melania Borit. More about the CfPM and him can be found on their website.

Juliette Rouchier is a cognitive economist based in a computer science lab, LAMSADE at Université Paris-Dauphine. She is interested primarily in trust in interaction, as an inter-individual phenomena or in a group performing collective action, and thus is focused on micro-social analysis mixing individual calculus, collective dynamics and social structure. She has been using ABM to study trust, reputation, information circulation, learning facing a complex world, social influence and self reinforcing phenomena. She has recently turned to the analysis of argumentation and rhetorics. Her main fields of application are food chains, environmental activism, change of practices. She is also interested in epistemological issues with ABM, role-playing games.

Martin Kunc is Professor of Management Science and Head of the Department of Decision Analytics and Risk. His research interests reside on the intersection between management science and behavioural science with emphasis on the use of analytics to support strategic decision making. He has experience in behavioural modelling for evaluating strategic decisions using principles of behavioural science and system dynamics.